Winter · The American journal of clinical nutrition 2014 · systematic review and meta-analysis of cohort studies · n=32 studies (197,940 individuals)

BMI and all-cause mortality in older adults: a meta-analysis.

Cited 801 times in the scientific literature.

Level 3 - non-randomized controlled study

Systematic review and meta-analysis of observational cohort studies.

PubMed 24452240 · doi:10.3945/ajcn.113.068122 · record verified 2026-08-27

What was done

A two-stage random-effects meta-analysis was performed on studies published from 1990 to 2013 evaluating the association between BMI and all-cause mortality in community-based adults aged 65 years and older. The analysis included 32 studies comprising 197,940 individuals with an average follow-up of 12 years.

What was found

Using a BMI of 23.0–23.9 kg/m2 as the reference category: - BMI 21.0–21.9 kg/m2: HR 1.12 (95% CI: 1.10, 1.13), a 12% increase in mortality risk. - BMI 20.0–20.9 kg/m2: HR 1.19 (95% CI: 1.17, 1.22), a 19% increase in mortality risk. - Being overweight was not associated with increased mortality. - Mortality risk began to increase only for BMI >33.0 kg/m2 (BMI 33.0–33.9 kg/m2: HR 1.08, 95% CI: 1.00, 1.15). - Results remained consistent after considering self-reported vs. measured data, adjusting for intermediary factors, and excluding early deaths or preexisting disease, with slight attenuation in never-smokers.

Why it matters

Standard adult BMI cutoffs do not accurately predict mortality in older adults, in whom being overweight is not associated with excess mortality whereas BMIs below 23.0 kg/m2 carry heightened risk.

Limits

The included studies are observational, leaving potential for residual confounding from subclinical illness, frailty, or unintentional weight loss. The abstract does not report cause-specific mortality, body composition metrics (lean vs. fat mass), or subgroup breakdowns by ethnicity.

Cited by